Chromosome inheritance ensures transmission of genetic and genomic information. Abnormal chromosome number (aneuploidy) and altered chromosome structure cause birth defects, reproductive abnormalities, and cancer. The centromere is the locus required for chromosome segregation and genome stability. Normal chromosomes typically have only one centromere, but, genome rearrangements associated with birth defects and cancer produce chromosomes in which two centromeres are physically linked. These dicentrics are not usually tolerated in most model organisms, as originally illustrated in maize by Barbara McClintock nearly 80 years ago. Paradoxically, dicentric chromosomes occur frequently in the general human population and are extremely stable during cell division. A major impediment in studying dicentric chromosome formation and fate in humans has been the absence of experimental systems. To circumvent this long-standing problem, we developed assays to experimentally create dicentric human chromosomes that molecularly mirror those that occur naturally and are biomedically relevant. We showed that in some of these de novo dicentrics, centromere inactivation occurred by partial centromere deletion. However, many of our engineered dicentric chromosomes, particularly dicentric X isochromosomes (dicXs), retain two active centromeres and are very stable. This finding appears to contradict McClintock's model of dicentric fate. In this proposal, we will build on our previous studies of dicentric human chromosomes by leveraging an inducible dicX assay system to explore molecular mechanisms governing stability of dicXs that maintain two active centromeres. We will focus on three major areas of investigation: 1) defining the molecular links between dicentric structure (i.e. inter-centromere distance) and centromere composition and kinetochore architecture; 2) testing the roles of alpha satellite genomic structure and transcription in dicentric stability, and 3) investigating mechanisms of centromere protein inheritance that result in varying centromere configurations on dicXs. Our work will place specific genomic and epigenetics events on the timeline of dicentric formation and stabilization by making use of a powerful chromosome engineering system that generates dicentric chromosomes that precisely model those that occur frequently in humans. These studies will also be critical for understanding dicentric formation and structure, refining long-established models of dicentric stability, and providing new molecular insights into inheritance of centromere function in humans.